Early Journal Content on JSTOR, Free to Anyone in the World
This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in
the world by JSTOR.
Known as the Early Journal Content, this set of works include research articles, news, letters, and other
writings published in more than 200 of the oldest leading academic journals. The works date from the
mid-seventeenth to the early twentieth centuries.
We encourage people to read and share the Early Journal Content openly and to tell others that this
resource exists. People may post this content online or redistribute in any way for non-commercial
Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early-
JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people
discover, use, and build upon a wide range of content through a powerful research and teaching
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please
AVERROES ON THE METAPHYSICS OF
AVERROES lived from 1126 to 1198. He was thus a con-
temporary of Maimonides (1 135-1204), and of Abelard
(1 079-1 142). He lived in Mohammedan Spain, and was the
last of the Arabian philosophers in that country. The governing
dynasty of the Almohades was not in favor of philosophical
studies, as leading to heresy and unbelief, and under this regime
a taboo was put on science and philosophy, their advocates and
students were proscribed and persecuted, and works dealing with
the forbidden subjects were confiscated and burned. As a result,
interest in the study, once so great as to influence the rest of
Europe and stimulate it to imitation and emulation, rapidly de-
clined. As a second result, the works of the Arabian philoso-
phers in the original Arabic are exceedingly rare, and a large part
of them lost, probably forever. Fortunately, translations were
made of them in Latin and Hebrew in the twelfth and thirteenth
centuries, and to these, especially the Hebrew, we are indebted
for the preservation of many works of the Arabian philosophers,
of which the originals are lost. This is especially true of Aver-
roes ; for he became, for reasons not quite certain, the favorite of
the Jews in Spain, Provence, and Italy, to the exclusion of Aris-
totle himself, of whom he was considered the commentator par
excellence ; and some of his works are extant in the Arabic lan-
guage, transcribed in Hebrew characters for the use of the Arabic
speaking Jews in Mohammedan countries. To the accident of
their transliteration they owe their escape from the Mohammedan
Besides works on medicine, jurisprudence, and astronomy, and
a treatment of some philosophical themes, Averroes is known
especially as a commentator of Aristotle, and in this department
his fame rests especially on his so-called "great" commentaries,
which he was the first to compose, and which have won for him
a mention in Dante's Inferno.
AVERROES AND ARISTOTLE. 4*7
As is well known, Averroes wrote commentaries to all the
works of Aristotle, and to some as many as three different kinds,
— "great," "middle," and "brief." The "great" commentary-
contains the text of Aristotle in full, and a detailed discussion of
the meaning. The " middle " commentary contains only the first
few words of the text of each paragraph, followed by a paraphrase
of the content of Aristotle's thought, closely following the order
and method of the original. In the " brief" commentary or resume,
or compendium, Averroes abandons the order of the original,
gives an exposition of the subject of the treatise in his own words
and by his own method, elucidates the problems under discussion
from the Aristotelian treatises bearing on the matter in hand, and
settles his account with his Arabian predecessors in the same field,
such as Alfarabi and Avicenna, particularly the latter. The short
commentary is thus a kind of independent work on the same sub-
ject as the Aristotelian treatise of the same name.
As was said before, the works of Averroes in the original
Arabic which are extant are very rare, and these form only a
small fraction of what he wrote. When Renan wrote his masterly
monograph " Averroes et I'Averroisme," which is still the only com-
plete work on the subject, though there is now new material for
elaborating and revising at least one section of that book, all that
was known to be extant in the original of the Aristotelian treatises
of Averroes were one manuscript in Florence, containing the
"middle" commentary on the Organon, the Rhetoric, and the
Poetics, and a second manuscript in the Escurial in Madrid con-
taining the commentary on the Psychology. Besides these he
knew of Arabic manuscripts in Hebrew characters of the com-
pendium of the Organon, the " middle " commentary of the
treatise On Generation and Corruption, on the Meteorologies, the
Psychology, and the compendium of the Parva Naturalia. These
were in Paris in the national library, and the Bodleian in Oxford
contained besides in the same characters the commentaries on the
De Ccelo, the Generation and Corruption, and the Meteorologies.
Since then a few other manuscripts have turned up, two in Ley-
den, one containing the " middle " commentary on the Organon,
Rhetoric, and Poetics, and the other his " great " commentary on
41 8 THE PHILOSOPHICAL REVIEW. [Vol. XVIII.
the Metaphysics. In Vol. XVIII of the Archiv fur Geschichte
der Philosophic, 1904-05, Derenbourg described another manu-
script of the Escurial library in Madrid containing commentaries on
some of the smaller physical treatises of Aristotle, and in Vol. XX
(1906—07) of the same periodical, Horten, of Bonn, the trans-
lator of Avicenna into German, called attention to the publication
in Cairo two or three years ago of the " middle " commentary on
the Metaphysics, under the editorship of Mustafa Al-Qabbani,
without any indication of the manuscript source of the edition. An
examination of the work proved to the present writer that it was
not the " middle " commentary, but the compendium, and a cor-
respondence elicited from the editor, Al Qabbani, the information
that the manuscript from which the edition was taken is in the
Khedivial library in Cairo, and contains other works of the same
author, and Professor Moritz, the director of the Khedivial library,
dates the manuscript about 700 A. H. (= 1322).
This little work, covering 85 closely printed large octavo
pages of small type of the Arabic edition, is one of the most im-
portant of Averroes's works, giving us as it does the commen-
tator's views on what he regards as the crowning point of phi-
losophy, the study of the separate intelligences, and of God.
Renan speaks of the importance of the treatise, and Munk gives
a brief statement of its contents, based upon the Latin translation,
or perhaps the Hebrew.
If any excuse is needed for presenting this subject to this asso-
ciation 1 now, I may say in the first place that no account of any
extent was given of it before, and that none of the writers on
Averroes, including the best known, Renan, Munk, and Stein-
schneider, saw the original Arabic, which became known only
about two years ago. It is unfortunate, however, that the Cairo
edition is not a good one, teeming with what are either mis-
prints or errors of the manuscript. The result is that while it helps
us decidedly to correct the Hebrew and Latin translations, and to
separate the numerous interpolations from the genuine text of
Averroes in these translations, we can by no means as yet dis-
pense with the latter, as, owing to the defective edition or manu-
1 This paper was prepared for the American Philosophical Association.
No. 4.] AVERROES AND ARISTOTLE. 419
script of the original, they help us in many cases to correct the
Arabic. Of the two translations the one that can least be dis-
pensed with is the Hebrew, and for the following reasons :
The Latin translation of Jacob Mantinus was made from the
Hebrew in the sixteenth century, and published in the Venice
Latin editions of Aristotle with the commentaries of Averroes. If
therefore we are sure of the Hebrew text we do not need the
Latin. The relation is not quite the same as between the Hebrew
and the Arabic. There is only one Arabic manuscript extant,
and, if we may judge from the Cairo edition, not an unusually
good one. It dates from the fourteenth century, whereas the
Hebrew translation was made by Moses Ibn Tibbon in 1258,
about sixty years after the death of Averroes. It is therefore
based upon an earlier manuscript than the one extant, and as
there are ten manuscripts extant of the Hebrew translation, it is
of no mean assistance to us in arriving at Averroes's correct text.
Another consideration should not escape us in considering the
relative value of a Hebrew and a Latin translation of an Arabic
original, apart from the dependence of one upon the other, as in
this special case. Arabic, being a Semitic tongue, can be so well
rendered literally into Hebrew, — and through many years of
translation from the former into the latter has been so rendered,
constituting a Hebrew philosophical style closely modelled after
the Arabic, — that one who is familiar with the two languages, and
with the subject, has no difficulty in reconstituting the Arabic
text from the Hebrew translation. That such a relation does not
hold between the Arabic and the Latin needs not my saying.
I have dwelt at length on this matter, because it is the fashion
in some quarters to belittle the Hebrew translations as worth-
less, and editors of Arabic philosophical texts are in the habit of
ignoring them, when they might be of great value. This applies
to men like J. Muller, Schmolders, Mehren, and others. Mehren
speaks of them as " de valeur suspecte," and Steinschneider
wonders whether the phrase is a cover for his inability to use
them. This is, I think, the secret in most cases. Sachau, the
Orientalist of the University of Berlin, also thinks, " man kann
mit ihnen gar nicht anfangen." Fausto Lasinio, on the other hand,
420 THE PHILOSOPHICAL REVIEW. [Vol. XVIII.
in his publication of excerpts from Averroes's logic, makes use also
of the Hebrew translations ; and Munk and Steinschneider, who
are really able to judge in this matter, are both of the opinion
that the Hebrew translations are of great value, and that a
knowledge of mediaeval Hebrew is more important in the study
of Arabian philosophy than a knowledge of Arabic itself.
The present study is based upon the Arabic text of the Cairo
edition above mentioned, upon a copy of the Hebrew translation
made by the present writer from seven manuscripts, and upon the
Latin translation as found in the Venice edition of Aristotle and
Averroes in Latin, 1573.
To judge from the Arabic text there is a number of interpola-
tions in the Hebrew translation, and the Latin always agrees with
the Hebrew. But as there is here and there an omission in the
Arabic, one is not quite sure that some of the apparent interpola-
tions in the translations are not rather omissions in the Arabic.
Averroes begins his treatise by dividing the sciences and arts
into three classes : (1) Theoretical, (2) practical, and (3) auxiliary
or logical. The aim of the theoretical is knowledge alone. In
the practical, knowledge is for the sake of action. The logical is
auxiliary to the other two.
The theoretical sciences are divided into two classes : Uni-
versal and Particular. The Universal considers the existent ab-
solutely and its essential attributes (pofjfieflrpcdTa xad' abvo). It
embraces (1) Dialectic, (2) Sophistic, and (3) Metaphysic. The
particular investigates the existent in a particular state. It em-
braces (1) Physics, which deals with changeable existence, and
(2) Mathematics, which deals with quantity abstracted from matter.
The three principal sciences, physics, mathematics and meta-
physics, correspond to the three kinds of existences : (1) Exist-
ence in matter (physics) ; (2) things existing in matter, but treated
apart from the latter (mathematics) ; (3) consists of two parts,
(a) principles existing absolutely not in matter (separate intelli-
gences, spirits of the spheres), and (&) Universals common to sen-
sibles and intelligibles, such as unity, plurality, actual, potential,
etc. The last two constitute the subject of metaphysics.
Of the universals he says in another place that as universals
No. 4-] AVERROES AND ARISTOTLE. 42 1
they have no extra-mental existence. They are not figments of
the brain ; they are not mere concepts ; they do exist objectively
in the concrete, but not as universals. The attribute of univer-
sality they do not acquire until they are apprehended by the
reason and the reason endows them with it. As objectively ex-
isting in the concrete they have a creative or productive power, —
man produces man, but not as universals. The mistake of the
Platonists, according to Averroes, is that they make the univer-
sals efficient causes.
Of the four causes, metaphysics deals especially with the formal
and final, and with the efficient also in a sense, i. e., not as pre-
ceding its effect in time, the sense in which it is used in physics.
He has in mind here God as the combination of the three causes
mentioned, formal, final, and efficient, but not preceding the uni-
verse in time, since motion is eternal. In physics the material
and moving causes alone are considered.
Averroes, it will be seen, divides the efficient cause into two
kinds, — the movens and the agens. The former belongs to
physics, the second to metaphysics. The first produces motion
only, in its effect, the second gives it form in virtue of which
motion takes place.
Metaphysics builds upon foundations laid in physics and mathe-
matics. From the former it accepts the idea of an immaterial
mover and shows in what way it moves. From mathematical
astronomy it accepts the number of movers, i. e., the number of
motions in the heavenly spheres, since each motion has a mover.
Metaphysics is thus divided into three parts. Part one deals
with sensibles qua existents, and all their genera, i. e., the ten
categories, and their aufifteftyxoTa xaffabvo. Part two considers
the principles of obaia, i. e., the separate intelligences, determines
the character of their existence, and relates them to their first
principle, who is God. It determines his attributes and actions,
the relations of other existences to him, and proves that he is the
ultimate perfection, first form and first agent. Part three inves-
tigates the subjects of the special sciences and refutes the errors
of former thinkers. By special sciences he means logic, physics,
422 THE PHILOSOPHICAL REVIEW. [Vol. XVIII.
Metaphysics occupies itself with this matter because the special
sciences do not verify their own principles. This is the province
of a universal science. Dialectics cannot do it, though it too is
a universal science, because it employs acknowledged but not
necessarily true propositions in its proofs. These may lead to
erroneous conclusions. Metaphysics alone uses true premises.
The third part, however, is not as essential to the science of meta-
physics as the first two. (It will appear from this that Averroes
does not agree with Herbert Spencer.)
These three parts of metaphysics Averroes divides into five
chapters, devoting the first three chapters to the first part of the
subject, the sensible qua existent, the fourth chapter to the sec-
ond part, viz., the separate intelligences and God. The fifth
chapter which was to have dealt with the third division of the
science, viz., the subjects of the special sciences, he seems never to
have written, as it is not found either in the Arabic or in the
translations. It is a pity that he did not write it, for under the
head of ' refutation of errors of former thinkers ' he would have
given us important historical material concerning the philosoph-
ical and theological sects of his day. On several occasions in
this treatise he refers to the views of the Mutakallimun, a school
of philosophical theologians of those days, and dismisses them with
a brief statement, deferring a more complete discussion of their
tenets to the fifth chapter, which he seems not to have written.
The purpose and value of metaphysics is the same as that of
the other theoretical sciences, viz., the perfection of the rational
soul. Metaphysics is more important in this respect because it
is the perfection of the other sciences, leading as it does to a
knowledge of existing things through their ultimate causes, and
verifying the principles of the other sciences.
In the order of teaching, metaphysics comes after physics,
since it makes use of certain principles laid down there, hence
called metaphysics. In essence, however, it comes before
physics, hence its other name, first philosophy.
The proofs in this science are of the kind called " signs " or
" indications " (oypzia of Aristotle), i. e., where we proceed from
what is better known to us to what is better known per se, or,
No. 4.] AVERROES AND ARISTOTLE. 423
which is the same thing, from a consequent to its antecedent.
The definition of this kind of proof, known as dalala in Arabic,
is given by Averroes in his compendium of logic as that in
which the middle term is not the cause of the conclusion, but
the cause of our inferring it, as for example when we prove that
the moon is spherical from the crescent shape of its light. The
crescent-shaped light is not the cause of its sphericity, but the con-
sequence. It is, however, the cause of our knowing that the
moon is spherical. It is really a kind of induction, as we should
The rest of the first chapter is devoted to definitions of terms
used in the science, closely modeled after the fifth book of Aris-
The second chapter investigates the ten categories, the sense
in which the term ' existent ' is applied to all of them, neither
synonymously nor homonymously, but per prius et posterius, or
as Aristotle calls it, npbt; sv, i. e., by virtue of their greater or
lesser participation in obo'ta, or reality.
The nine categories of accident are all dependent upon the first.
It is independent of them. The definition is found first in ouaia.
In the other categories, if it exists at all, it is secondary.
The essences or universal concepts of things are identical with
the things themselves, else knowledge would be impossible. The
Platonists, who place the concepts outside of the concrete, either
make knowledge impossible, or they require another set of con-
cepts to understand the first, and so on ad infinitum.
Neither the form nor the matter is generated or destroyed.
Hence it follows that the definition is neither generated nor
The universals as universals exist in the mind only. The uni-
versal as universals are not the essences of things. Sensible
things are composite. This is proved by the fact that we ask the
question, Why ? This cannot be asked of the simple. We can only
ask, Whyis A ? Why is B ? The answer to the question, ' why,' is
any one of the four causes. Though composite, the concrete things
do not contain the elements of which they are made in actu, or
they would be nothing else but the elements, which is not true-
424 THE PHILOSOPHICAL REVIEW. [Vol. XVIII.
There is in the compound a something outside of the elements
which makes it what it is, and that is the Form.
In the definition, the genus corresponds to the matter, the dif-
ference to the form.
Matter is that which is in potentia. Form is actuality. Geo-
metrical figures, too, have a kind of matter, and hence have a
Things have a two-fold existence, — sensible and intelligible.
Intelligible existence is sensible existence as known.
If a definition has parts, the compound alone has a definition.
Matter, form, and simple things in general have no definition.
Multiplicity is due to matter, unity to form. Though a composite
of matter and form, and possessed of a definition containing parts,
the concrete is a unit, because the combination of matter and
form means the realization of the potential.
Chapter III deals with the aufiftePyxora xaff abro of the ten
categories, such as actual and potential, the one and the many,
and the " Contraries," and the finiteness of the four causes, and
their relations to one another. Time will not allow me to enter
into details, and I must proceed to a very brief sketch of
the fourth chapter.
All that preceded has led up to this last chapter, in which Aver-
roes proves from the eternity of motion the existence of eternal
immaterial movers existing actually, and a single principle, which
is God, existing as the cause of the latter and of the universe.
His proofs are based upon the twelfth book of Aristotle's Met-
aphysics, except that he develops in detail what Aristotle has left
in broad outline. I shall therefore assume Aristotle's scheme as
known and say a few words about that which is peculiar to
In deducing the number of motions and of movers of the
celestial sphere (55), Aristotle prefaced his discussion by dis-
claiming any dogmatism in the matter, saying that the subject is
one of probability, not of certainty, and that he was willing to
adopt the conclusions of the astronomers of his day, modifying
them as he saw fit, and leaving an open door for later revision
consequent upon better knowledge. He thus adopts Callippus's
No. 4-] AVERROES AND ARISTOTLE. 425
correction of Eudoxus as to the number of the spheres, which he
further revises by the addition of the counteracting spheres.
Averroes, who was also somewhat of an astronomer, takes this
hint from Aristotle, ignores the schemes of Eudoxus, Callippus,
and Aristotle himself, and adopts provisionally again, the system
of Ptolemy as the best that was known at the time, saying that
the whole matter was still uncertain, owing to the difficulty of the
subject and the imperfection of the instruments. He thus adopts
the view that there are 38 motions, 5 each in the upper three
planets, Saturn, Jupiter and Mars, 5 in the moon, 8 in Mercury,
7 in Venus, and one in the sun, on the hypothesis of an eccentric
and not of an epicycle, for that would make two. This is the
number (38) if we suppose that the mover responsible for the
diurnal motion is the same for all. If a different one is required
for each of the 7 planets, there will be 45 movers, or separate
intelligences, in all.
It is interesting to note here a suggestion of Averroes, that it is
possible that instead of Saturn receiving its motion from the
sphere of the fixed stars and the others from Saturn in the tradi-
tional order, the sun comes next to the fixed stars, then Saturn
and the rest in order. This idea commends itself to him from
the observation of the dependence of the other planets upon the
motions of the sun in speed and proximity. Would it be too
much to call this an adumbration of the heliocentric theory ? It
would seem to have been original with Averroes, as he does not
mention any authority for the statement.
Aristotle does not discuss in detail the functions and characters
of the movers of the spheres, or their relations to each other and
to the primum movens, or God. Of the latter he says that he is
pure reason contemplating himself eternally, and that the heavenly
bodies move eternally, not because he moves them directly, but
because they are prompted by a love and a desire for God.
Averroes supplies the deficiency.
The movers of the spheres are pure intellects, the upper being
the cause of the lower. The heavenly bodies are endowed with
life and reason, and their motions are a result of conception and
desire. They conceive the good, which is their perfection, and
426 THE PHILOSOPHICAL REVIEW. [Vol. XVIII.
desire to become like unto it. Inasmuch as motion is better than
rest, since motion is life, they are constantly in motion. The
movers not only move the heavenly bodies, but they give them
their forms in virtue of which they are what they are, and hence
they are thus agentes, in a sense. The forms of the heavenly
bodies are what they conceive of the movers above them.
As to the movers, or principles, they are pure intellect and their
function is knowing, conceiving. They are immaterial and im-
movable. Judging from our own intellect, which knows itself,
we infer that each of the separate intelligences knows itself. To
know itself, it must know that upon which it depends, hence it
must know its cause, which is its perfection, i. ^.,the mover next
above it, though not in the same way in which the latter knows
itself, or they would be identical. The cause, however, does not
know its effect, which is inferior to it, hence the last mover, or
God, having no cause higher than himself and not knowing the
lower movers, knows himself only.
If Averroes stopped here he would be denying all knowledge in
God of things below, and hence all Providence. He is not ready
to do either. It is absurd, he says, that anything should emanate
from a knower qua knower without the latter knowing it, and
an observation of nature is all that is needed to recognize Provi-
dence. He therefore compromises on his previous deductions
and argues again from the human mind to prove that the upper
intelligences do have a knowledge of the lower. Since our intel-
lect, he says, is nothing more than a conception of the order and
method of this world and its parts in reference to its causes,
proximate and remote, the essence of the intellect producing ours
(one of the movers of the lunar sphere) cannot be different, except
that it comprehends the same thing in a superior manner. The
same holds true of the intelligence next above this, and so on to
the first intelligence, or God. It follows then that God knows
the same things as we know, but in a superior manner. These
two deductions are both true. God's knowledge of himself is
identical with his knowledge of the Universe. But his knowledge
is not our knowledge.
Upon this compromise Averroes bases his theodicy. God is
No. 4-] AVERROES AND ARISTOTLE. A^7
not responsible for evil, for evil is a concomitant of matter, and
that God does not know. Not to know some things is better than
to know them, and argues no defect or imperfection, rather the
contrary. To make God know the particular and save him from
responsibility for evil by saying that good and evil are only in
relation to us, that to God they do not exist, is a dangerous
It would seem as if all great men who do epoch-making work
require as a stimulus, in addition to the cause of truth, an
embodiment of a deviation from it in the shape of a personal
opponent against whom they may sharpen their wit, thus rising
to greater heights than the smaller men about them. Aristotle
had his Plato, Abelard his William of Champeaux, Kant his
Wolff, and Averroes his Avicenna. He finds no less than eight
occasions in this little book to signify his disagreement with the
latter, and in one instance declares that Avicenna can never be
relied upon when he goes a-hunting after original views.
The Mutakallimun, or Arabian Scholastics, were another bug-
bear to his rigid Aristotelianism, and he does not treat them with
great gentleness. " They do not carry on their discussion," he
says, " by means of syllogisms composed of two premises, nor
do they make use of essential predicates." This condemnation
of their unphilosophical, because unlogical, method is only
equalled by another in which he says of them : " they do not hold
these views (such as the denial of causation and the like) because
they are led to them by investigation, but in order to verify by
their means opinions about which their minds are made up in ad-
vance, and refute principles opposed to them." Maimonides
speaks of them almost in the same words. We should have had
more information about their system if Averroes had written the
fifth chapter which he promises on the subjects of the special
Though the reputation of Averroes for orthodoxy did not rank
high, still he takes the opportunity to quote the Koran in a few
instances in confirmation of his philosophical views, such as the
unity of God, his knowledge of things in the world, and the im-
portance of knowing the human mind before studying to under-
stand the nature of God.
428 THE PHILOSOPHICAL REVIEW.
Many and varied have been the opinions regarding Averroes as
a commentator and exponent of Aristotle. He was once glorified
by Jews and Christians alike as the commentator far excellence.
He was as much depreciated later in the time of the Renaissance,
and hated by a man like Petrarch with almost a passionate hatred.
Then he was simply ignored and forgotten. In Solomon Munk
he has found a defender who thinks we may still consult him with
profit at the present day. It will be near the truth, I think, if
we recognize that for his day he was the best exponent of Aris-
totle, better than any of his predecessors ; that considering he was
twice removed from the original text, the Syriac translations
having stood between the Greek and the Arabic, he was as effi-
cient and penetrating a commentator as can be imagined ; and
while there is no need of consulting Averroes now when we have
Aristotle's text and the Greek commentators, and Bonitz's index
and the other works of Bonitz, and Trendelenburg, and Waitz,
and Zeller, and others, we may study him as one of the sources
of mediaeval philosophy, who will help us to understand men like
Albertus Magnus, Thomas Aquinas, William Occam among the
scholastics, and Gersonides, Falaquera, and Caspi among the
Jewish philosophers of the later middle age.
University of Pennsylvania.